Feedforward Finite Control Set Model Predictive Position Control of PMSM

Abstract

This paper presents a predictive controller that simultaneously controls the position, speed and the current of PMSM. The controller is based on finite control set model predictive control approach. Motion profile generator is used for the generation of reference signals. Signals generated there enter into the controller and act as feedforward signals that are evaluated inside of a cost function with other predictions. Measured position is used as a feedback and the speed and load torque are estimated by Kalman filter. The control is formulated in the dq coordinate frame. Simulation results show that the performance of the proposed position controller is significantly improved by using the proposed feedforward terms in the cost function when comparing to the cost function without these terms. The controller with improved cost function was also experimentally verified with Texas Instruments DSP.